The present invention relates to a CT (Computer Tomography) imaging method and an X-ray CT apparatus, and more particularly to a CT imaging method and an X-ray CT apparatus capable of producing a CT image in a short time.
FIG. 5 is a diagram explaining the conventional CT imaging method. For the expedience of explanation, one frame of image is assumed to be composed by eight pieces of image data A11 through A42. The X-ray sensor array in use is assumed to have four channels, of which the 1st-channel sensor produces image data A11 and A12, the 2nd-channel sensor produces image data A21 and A22, the 3rd-channel sensor produces image data A31 and A32, and the 4th-channel sensor produces image data A41 and A42. The X-ray sensors of 1st through 4th channels are provided with modification data B1, B2, B3 and B4, respectively. The image data pairs A11 and A22, A21 and A32, A31 and A42, and A41 and A12 are each derived from parallel X-ray transmission paths.
The X-ray CT apparatus has a central processor, which is assumed to include a master processor MP and four slave processors SP1 through SP4.
The master processor MP groups the image data based on their originating X-ray sensors, and distributes the resulting paired data to the slave processors SP1 through SP4. At the same time, the master processor MP sends the modification data corresponding to the X-ray sensors to the slave processors SP1 through SP4.
Specifically, the first slave processor SP1 is supplied with data pair A11 and A12 produced by the 1st-channel sensor and the modification data B1 corresponding to that sensor. The second slave processor SP2 is supplied with data pair A21 and A22 produced by the 2nd-channel sensor and the modification data B2 corresponding to that sensor. The third slave processor SP3 is supplied with data pair A31 and A32 produced by the 3rd-channel sensor and the modification data B3 corresponding to that sensor. The fourth slave processor SP4 is supplied with data pair A41 and A42 produced by the 4th-channel sensor and the modification data B4 corresponding to that sensor.
Since a data pair supplied to each slave processor have fan-out X-ray transmission paths, these image data will be called "fan-view data".
The slave processors SP1 through SP4 implement the first pre-processing for the supplied data pair A11 and A12 through data pair A41 and A42 by using the supplied modification data B1 through B4, thereby producing a new data pair a11 and a12 through data pair a41 and a42.
The master processor MP collects the data pair a11 and a12 through data pair a41 and a42 produced by the slave processors SP1 through SP4, groups the data into data pairs each derived from parallel X-ray transmission paths, and distributes the paired data to the slave processors SP1 through SP4.
Specifically, the image data is grouped into data pair a11 and a22, data pair a21 and a32, data pair a31 and a42, and data pair a41 and a12, and these data pairs are supplied to the first, second, third and fourth slave processors SP1, SP2, SP3 and SP4, respectively. Since a data pair supplied to each slave processor have parallel X-ray transmission paths, these image data will be called "para-view data". The above-mentioned data grouping process from fan-view data into para-view data will be called "fan-para data conversion".
The slave processors SP1 through SP4 implement the second pre-processing for the supplied data pair a11 and a22 through data pair a41 and a12, thereby producing a new data pair .alpha.11 and .alpha.22 through data pair .alpha.41 and .alpha.12.
The slave processors SP1 through SP4 further implement the FBP (Filtered Back Projection) process for the data pair .alpha.11 and .alpha.22 through data pair .alpha.41 and .alpha.12, thereby producing new data i1 through i4. The resulting back-projected data i1 through i4 will be called "BP data".
The master processor MP collects the BP data i1 through i4 produced by the slave processors SP1 through SP4, and composes the data to produce a CT image I.
As described above, the conventional scheme is designed to distribute the fan-view data to the slave processors SP1 through SP4 to carry out the first pre-processing. It is convenient in that the slave processors SP1 through SP4 are merely supplied with the modification data B1 through B4, respectively.
However, it suffers an extended processing time due to the data transfer for the collection of data pair a11 and a12 through data pair a41 and a42 resulting from the first pre-processing to the master processor MP and the data transfer for the distribution of para-view data pair a11 and a22 through data pair a41 and a12 to the slave processors SP1 through SP4.